Employment patterns of older persons in the United States have changed dramatically in the last several decades, marked by two important trends. The first is the trend towards earlier retirement that dominated much of the last century, but currently shows signs of reversing (Costa 1998). The second is the decreasing stability of employer-employee relationships, in the form of increased layoffs, to which older workers are disproportionately exposed (e.g. Osterman 1999;Stevens 2001). The purpose of this project is to provide evidence on the health and health care utilization implications of these trends, focusing separately on the effect of voluntary reductions in latter-life employment and the effect of involuntary disruptions in latter-life employment. The existing empirical literature provides limited and somewhat conflicting evidence regarding these effects. Problems of omitted variable bias plague most analyses, as older workers who exit employment, either voluntarily or involuntarily, are likely different from others across unobserved dimensions like health status. The current application uses a unique Norwegian population-based dataset to overcome these limitations and address the following aims: (1) to estimate the effects of involuntary work disruptions on the health and health care utilization of older workers;and (2) to estimate the effects of earlier retirement, induced through earlier availability of retirement benefits, on the health and health care utilization of older workers. For each analysis, the project will utilize a "natural experiment" approach to address problems of omitted variable bias. To study the effects of involuntary work disruptions (Aim 1), outcomes of older workers will be compared across workers employed in plants that subsequently closed and similar workers in stable and growing plants. To study the effects of voluntary reductions in latter-life employment (Aim 2), the analysis exploits changes in an early retirement program that induced earlier retirements among workers eligible for the program. Using a difference-in-differences framework, trends in the health and health care utilization of eligible older workers (i.e. before and after the change) will be compared to trends among ineligible older workers to draw credible causal inferences. The project will utilize an impressive array of longitudinal Norwegian data, linking individual and household characteristics, plant-of-employment characteristics, mortality outcomes, and health care utilization data (with dates and diagnoses) pertaining to all hospitalizations and specialist services obtained through the Norwegian health care system. While the use of Norwegian data raises issues of generalizability to the American population, the mechanisms through which work and employment disruptions affect health are likely to be similar, and the data allow for a more complete accounting of health and health care utilization effects than could be accomplished using existing U.S. data sources. PUBLIC HEALTH RELEVANCE: Recent changes in retirement behavior and in the security of employment relationships potentially have important implications for the health and wellbeing of older workers, as well as for the financing of health care for the growing population of older persons. However, the effects of involuntary work disruptions (through layoffs) and voluntary reductions in latter-life employment (through earlier versus later retirement) on health and health care utilization remain largely unknown. Credible causal estimates of these effects can help guide appropriate policy responses to these ongoing labor market trends, as well as inform policymakers of the health, welfare and fiscal implications of proposals to encourage latter-life employment.